Impact of Homogeneous and Heterogeneous Content Store Size Allocation on Cache Placement Policy For Named Data Network
Kemal Fadhlurrahman Pribadi, Ridha Muldina Negara, Leanna Vidya Yovita · 2022
Named Data Networking was developed to cache efficiently, which can be done with a cache policy strategy by allocating homogeneous and heterogeneous content stores. In a homogeneous cache, data packets stored on network routers have the same size as the content store. Meanwhile, heterogeneous cache stores data packets on each router in the network having different content store sizes. There are several optimization techniques based on replacement algorithms and placement algorithms. The replacement algorithm is Least Frequently Used (LFU), where the placement algorithm includes Leave Copy Everywhere (LCE), Core Caching, Edge caching, and Probability. This paper analyzes the policy of placing cache content stores (CS) with homogeneous and heterogeneous sizes and using replacement and cache placement algorithms. System performance will be seen from the cache hit ratio and delay. This study conducted a simulation to test the cache placement scheme on an NDN network. Tests were performed by changing the content store, resizing the frequency of interest, and changing the Zipf size exponentially. LCE cache deployment performs better than the other three cache deployments. The cache placement probability performs better than LCE when the scheme changes the desired frequency at 200 and 500 frequencies. That proves that cache placement Probability can perform better when the network is busy.